Accuracy study of colloidal gold-based rapid test reagents for Chinese food

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Objective To evaluate the accuracy of food rapid test reagents based on colloidal gold technology for the detection of chloramphenicol, enrofloxacin and sulfonamides in livestock, poultry and aquatic products, and to provide a reference for their practical application. Methods Rapid test reagents from different manufacturers were used to detect the above targets in pork, chicken, fish, shrimp, and other samples. The specificity, linearity and sensitivity of the reagents were analysed by setting up a multi-concentration gradient combined with the interference experiment of flavouring matrix. Results Reagents D and E showed strong anti-interference ability and specificity for chloramphenicol, though reagent E yielded false positives at 0.05 µg/kg. For enrofloxacin, polynomial fitting correlation coefficients (0.9991, 0.9966, 0.9993) outperformed linear regression. Compared to reagents using sulfadiazine or sulfamethoxazole as the benchmark, those based on sulfamethazine demonstrated superior accuracy. Reagent B achieved an overall accuracy rate of 96.6%. In contrast, reagents using sulfamethoxazole as the benchmark showed no response to sulfadiazine, sulfamerazine, or sulfamethazine, resulting in a false negative rate as high as 100%. Conclusion The colloidal gold detection technology is effective in screening for chloramphenicol in seasoned meat and fish samples; for enrofloxacin, it exhibits a good linear relationship and is suitable for preliminary screening in the circulation chain. The sensitivity for sulfonamides depends heavily on the benchmark substance used in reagent design. We recommend selecting rapid tests developed with a structurally representative benchmark, like sulfamethazine, and validating them across matrices before deployment. Confirm with traditional methods if needed.

Article activity feed